Social media data: advanced processing and modeling

Type: Virtual Paper
Sponsor Groups:
Poster #:
Day: 4/9/2021
Start / End Time: 11:10 AM / 12:25 PM (PST)
Room: Virtual 25
Organizers: Oleksandr Karasov, Evelyn Uuemaa, Tiit Tammaru
Chairs: Oleksandr Karasov

Call for Submissions

This session welcomes the abstracts on any "real world" social media data applications, which contribute to a deeper understanding, mapping, or modeling of environmental, social or economic processes. Usually, different disciplines use field-specific tools and methodologies (e.g., natural language processing, machine learning, color analysis). We will discuss the state-of-the-art in quantitative and qualitative social media data research to embrace and further integrate novel tools and methodologies, directly or indirectly influencing various aspects of sustainability. We would like participants to specifically address the problems of social media data applications, which participants have successfully solved, or looking for a solution.


Social media data comprise a wealth of user-generated content enabling the high-fidelity insights into the environmental, social and economic dimensions of sustainability. Richness and diversity of social media data has recently boosted socioeconomic research, tourism and environmental protection, digital anthropology and cultural analytics, urban studies. A recent radical decline in offline people’s activities and mobility during the COVID-19 pandemic has exacerbated the environmental and socioeconomic inequalities and increased the part of time, which people spend online.

However, a rapid development in this field imposes new constraints and challenges, including, but not limited to: changing social media sites and their data policies, non-uniform spatial component in the user-generated content among websites, complexity of content analysis of texts, images and videos, insufficient recognition of synergies between online and offline data sources, problems of uncertainty assessment in social media data-based models. These problems and respective field-specific solutions increase the gap between the social media data applications among disciplines and social media sites.

This session aims to deconstruct the existing landscape of social media applications to mutually enrich the existing methodological pathways and find ways for more effective operationalization of social media data. In this session, a lively discussion among several social media experts will unveil not only the problems, which social media help to solve but also the most relevant issues, social media scholars face with.


Type Details Minutes Start Time
Introduction Oleksandr Karasov University of Tartu 1 11:10 AM
Presenter Mark Wilson*, Michigan State University, Eva Kassens-Noor, Michigan State University, Travis Decaminada, Michigan State University, Cornelius Darcy, Michigan State University, Smart City Tweets: The Role of Influencers on Technology Messaging 15 11:11 AM
Presenter Ruihong Huang*, Northern Arizona University, Discovering National Parks Visitor Activities Using Geotagged Photos 15 11:26 AM
Presenter Volodymyr Mihunov*, Louisiana State University, Lei Zou, Texas A&M, Nina Lam, LSU, Zheye Wang, LSU, Kejin Wang, LSU, Dylan Govender, LSU, Analyzing the use of Twitter in hurricanes and floods: common approaches, lessons learned, and emerging trends 15 11:41 AM
Presenter Avijit Sarkar*, University of Redlands, James Pick, University of Redlands, Tridev Raut, University of Redlands, Purposeful Use of Social Media in U.S. Counties: A Multivariate and Geospatial Perspective 15 11:56 AM
Discussant Oleksandr Karasov University of Tartu 14 12:11 PM

To access contact information login